There's a category of software that every large enterprise needs, that the big vendors price out of reach and major cloud providers like AWS and Azure won't build because it would impact their pro-serve revenue.
I've spent 10 years finding that category. Now I'm building inside it.
The gap nobody talks about
Amazon Connect is genuinely good contact center infrastructure. It's cheap. It's elastic, it integrates natively with the rest of AWS, and it's taken significant market share from legacy vendors like Genesys and Avaya over the last five years.
But "Contact Center Infrastructure" and "Contact Center Intelligence" are different problems.
Connect handles the calls. It doesn't tell you whether you're compliant. It doesn't surface which agents are struggling before your CSAT scores crater. It doesn't give you a single coherent view across the five AWS services that together hold everything you'd want to know about your contact center operations.
Those five services — Contact Trace Records, Agent Events, Amazon Lex, Amazon Bedrock, and Contact Lens — each tell you part of the story. CTRs give you call metadata. Agent Events give you desktop activity. Lex tells you about bot interactions. Bedrock handles AI-powered conversations. Contact Lens surfaces sentiment and compliance alerts. None of them talk to each other in a useful way out of the box.
So the intelligence layer — the thing that ties it all together — is either done manually by your compliance and QA teams, or it's done by expensive third-party platforms that require six-figure contracts and six-month implementations.
Why neither answer works
The manual approach doesn't scale. If you're running 100+ agents across multiple queues, having a compliance team manually audit call recordings for PCI-DSS scope or HIPAA data handling isn't viable — it's just risk accumulation with extra steps.
The third-party platform approach works, but only for the customers it was designed for: large enterprises that can absorb the cost and the implementation timeline. Verint, NICE, Calabrio — these are $500K+ annual contracts with 6-month onboarding. They are not built for a 300-seat contact center running on AWS.
And AWS itself? AWS makes money when enterprises hire AWS Professional Services to build the intelligence layer themselves. There's a structural reason they haven't productized this — it would cannibalize the engagement revenue.
The mid-market is caught in the middle
Mid-market companies — the ones running 50 to 500 seats on Amazon Connect — are the fastest-growing segment of Connect's customer base. They chose Connect precisely because it's cost-effective. And then they discover that operating it compliantly and intelligently requires either expensive people or expensive software.
PCI-DSS v4.0 and HIPAA compliance for a contact center involves 18+ discrete controls: call recording encryption, data retention windows, agent authentication, cardholder data handling, right-to-erasure workflows. These aren't optional. They're audit requirements.
Most mid-market teams handle this with spreadsheets and quarterly reviews. That's not compliance. That's hope.
What the intelligence layer actually needs to do
It needs to be deployable in hours, not months — Terraform-native, no professional services required. Compliant by default, with PCI-DSS v4.0 and HIPAA controls built in, not bolted on. Unified across the AWS services you're already using, not another data silo. And priced for the mid-market, not a six-figure contract.
That's what we're building at Lokha Labs.
ConnectAudit covers the compliance layer: 18 automated controls, a QuickSight dashboard, and zero manual audit prep. Contact 360 handles the intelligence layer: a unified view across CTR, Agent Events, Lex, Bedrock, and Contact Lens — five data sources in one query, MCP-native.
If you're running Amazon Connect and doing compliance and QA the hard way, we're accepting design partners now.